Growing a winner

Summary: Software Innovation

This is a summary with links to my posts on software innovation. It includes posts on barriers to innovation and how to grow a winner.

Software is taking over the world. The pace and scope of the transformation of human activity has no precedent. People often assume that this is the result of fierce innovation in software. While brand-new software is built every day, the actual innovation is the result of computer innovation – which does indeed proceed at an unprecedented pace.

The fact is that innovation in software is incredibly simple. It rarely involves fancy stuff like AI, but mostly figuring out the best way to accomplish things and getting the computer to do as much of the work as possible.

Here are some of the most important patterns of the small groups that innovate their way to success in software.

The general impression is that software is all about innovation and rapid evolution. Programmers are on the front lines, constantly making new things. Sadly, this general impression, which is shared by most programmers, doesn't hold up under examination.

People love to brag about the software innovations they’ve invented. The fact is, most of the fundamental innovations in software are proven and in place; they’re ignored by practically everyone, including the experts.

This holds true even when you look at a narrow field of application such as financial technology.

An amazing fraction of what appears to be innovations are little but taking advances that are proven in a narrow domain and applying them to a new one. Here is the story of an algorithm that was standard practice in oil refinery operation over 50 years ago that, decade by decade, is still crawling its way into new domains.

Here is an example of a truly beneficial innovation proven over 50 years ago that is still not used in medical imaging.

It’s not just fancy algorithms that are proven and waiting for application in new domains. It’s simple things like production human data entry, where widely proven “heads down” methods are at least 5 times more efficient than what is normally done.

Here are details of the advantages of “heads down” data entry and how it was ignored at a huge project at Sallie Mae.

One of the common ways to ignore a major innovation such as “heads down” data entry is to concentrate on a method that is highly fashionable, even though it doesn’t do much good.

Given this, it’s all the more amazing that companies have Chief Innovation Officers whose job is usually to “foster innovation.” Heh.

Barriers and Resistance to Innovation

Software innovation faces huge barriers from people and established practice, like innovation in medicine, where a cure for scurvy was proven in practice for decades before the authorities grudgingly accepted it.

Innovation has been strongly resisted for a long time.

Have you heard the story of Samuel Pierpont Langley, the renowned expert on manned flight? He’s a case study in how experts prevent innovation.

Major advisory institutions reflect common thinking and prevent their customers from "making mistakes" with innovative companies.

Here is the story of how the British military resisted a huge innovation for combating submarines in World War 2.

Even the experts in entrepreneurship resist innovation. Amazing. Experts!

Perhaps you think that the big tech companies are great at innovation? When you look at how they actually innovate, they don’t look so great.

Often important innovations don’t require software at all, which doesn’t seem to stop people spending loads on money on exotic software.

Healthcare is a rich source of examples of how to screw up and fail to take advantage of obvious, long-overdue “innovations” like electronic medical records.

While experts and big companies put up active resistance to innovation, regulations are an important source of passive resistance.

The huge cost of medical imaging systems is a clear example of how regulations prevent innovation and keep costs high.

When you look into sample sets of regulations, you see how onerous they are and how they hamstring innovation.

There’s a clear path to eliminating regulatory resistance while making the enforcement power of the regulations even stronger, by shifting from massive how-type regulations to simple but effective what-type ones.

When you look at how innovation works in practice, it's hard to avoid thinking that there's a conspiracy to prevent innovation. It's mostly not a conscious effort on the part of the conspirators, but it has the same net effect.

Growing the Innovation to Success

Once you’ve got your software company up and running, there are strategic moves that will keep you on the track to success.

Would you like to follow Facebook’s growth path? Their success wasn’t about great software development. It was due to a classic product/company growth strategy.

Your strategy may be good, but unless you build applications that can be changed quickly, you’ll lose the race.

When you’re moving quickly, classically-trained programmers may start whining about the growing amount of “technical debt” and how important it is to pay it down. Here’s how to do it.

If you follow classic software development methods instead of fast ones, the chances are you could be hit with a big surprise at the end of the project.

There is a way to avoid disasters of this kind. It's practicing wartime software development, optimizing your methods for speed instead of meeting expectations.

When you’ve built a successful narrow-focus software company, how can you grow it further?

The Book

Here are highlights of winning growth strategies and the book with more stories and details.

Here are highlights of the companies used as examples in the book.